Job Description

• Masters in computer science, data science, electrical engineering, applied mathematics, or a related field with 3+ years of work experience
• Demonstrated leadership abilities
• Excellent written and oral communication skills

Job summary The Alexa Automatic Speech Recognition (ASR) science team in India is rapidly growing. The team is responsible for advancing Alexa’s core ASR technology (in collaboration with the EU and US counterparts) and building highly accurate ASR production models in multiple languages, such as English dialects (UK, AU, NZ, IN) and Indic (Hindi etc.) languages. The challenge is ASR models must generalize to various device form factors (smart speaker, multi-media devices, etc.), acoustic conditions (far-field, close-talk, mobile, and noisy), and content (natural conversations etc.). As part of the expansion, the team will be responsible for complex ML technologies such as (i) learning without relying on human transcribed data by leveraging abundant unlabelled data via semi-supervised learning (SSL) and self-supervised learning, and customer feedback signals through weak-supervision (WS) (ii) Online Learning and Lifelong Learning, which requires a major overhaul of the learning infrastructure and ML algorithms as data will not be stored for training (iii) Privacy-preserving Federated-Learning, which enables training where audio does not even leave the devices. (iv) multilingual ASR technology, where a single ASR model will serve multiple languages (e.g., Indic languages). Apart from working on cutting-edge ML technology, you will have the opportunity to work on some exciting new products and also publish your work in top-tier conferences.

• Ph.D. in computer science, data science, electrical engineering, applied mathematics, or a related field
• Solid research track record with peer-reviewed publications in relevant areas
• Scientific thinking and the ability to invent, a track record of thought leadership and contributions that have advanced the field
• Hands-on experience in machine learning and programming
• Experience in professional software engineering best practices
• Comprehensive knowledge of speech and language technology, including speech recognition and natural language understanding